New high power pulsed transmitters are an essential feature of most new search radar and track radar designs. Because these transmitters usually incorporate the most recent advances in component technology and because these components are generally subject to great operating stresses, transmitter development frequently presents one of the most troublesome areas in overall radar design. Such parameters as gain, bandwidth, output power and broadband noise are generally emphasized as measures of transmitter performance. Equally important narrowband parameters such as intrapulse noise and pulse-to-pulse coherence are relegated to a lower level. These latter two performance indicators, however, become increasingly important transmitter figures of merit as the radar system design becomes more critical.
The trend in radar design is toward coherent systems with complex modulations using such techniques as pulse compression, pulse doppler, and moving target indicator (MTI). These techniques require transmitter characterization in terms of its contribution to the system error budget when tested with the system waveform.
Two basic approaches have been used in the past to determine the contribution of the transmitter to the system signal processing error budget. The first of these requires the measurement of transmitter intrapulse noise by spectral analysis and the estimation of error budget contribution from these data. The second approach requires the complete radar system and the knowledge of the error budget contributions of the remainder of this system.
There are two basic approaches that have been used for performing a spectral analysis of the transmitter output. The most straightforward approach is analysis at the carrier frequency using an RF spectrum analyzer. The performance limitation of the test equipment is the greatest problem with this approach. Due to the low frequency resolution of RF spectrum analyzers, this RF approach has no value for intrapulse noise measurements, although it has been used to examine wideband transmitter noise effects.
The second spectral analysis approach both improves the resolution and provides isolation of the item under test from input signal instabilities. This procedure requires that the item, either a signal amplifier or a sequence of series gain stages, be inserted into a phase bridge as described in the article "The Measurement of Near-Carrier Noise in Microwave Amplifiers," by Klous H. Sann (IEEE Transactions on Microwave Theory and Techniques, Volume MTT-16, Sept. 9, 1968). The phase bridge approach reasonably satisfied dynamic range and resolution requirements but does not give an adequate system-oriented characterization of transmitter capability due to the difficulties encountered in translating the AM and PM noise spectra into system performance figures of merit such as MTI cancellation and due to the type of averaging required by the spectrum analysis. The long-term averaging requirement of the phase bridge approach obscures the pulse-to-pulse stability characteristics which are of prime importance, for example, to the MTI system.